4 ms·

The idea would be great if it could scale to 100B weights. In the loop training is cool - forward prop using the neuromorphic hardware, network update using Pyt

The idea would be great if it could scale to 100B weights. In the loop training is cool - forward prop using the neuromorphic hardware, network update using Pytorch. But for now, they only get 98.7% on MNIST, state of the art is 99.91%, a 14x lower error rate. You have to try hard to get under 99%, just take a look. Maybe it's because they can't use backprop. Backprop is so powerful it's hard to beat it. https://paperswithcode.com/sota/image-classification-on-mnist?metric=Accuracy https://paperswithcode.com/sota/image-classification-on-mnis... https://www.kaggle.com/c/digit-recognizer/leaderboard https://www.kaggle.com/c/digit-recognizer/leaderboard

billlli 5y ago
One of the authors here… Scaling certainly is one of the next big challenges, the current network sizes severely limit us in our inference performance. Just to clear things up: Our circuits were actually trained via backprop. This is what allowed us to reach performance levels very close to equivalently sized but simulated SNNs (and even rather close to the accuracy of ANNs of the same size).